Weekends-off efavirenz-based antiretroviral therapy in HIV-infected children, adolescents and young adults (BREATHER): Extended follow-up results of a randomised, open-label, non-inferiority trial

Background: Weekends off antiretroviral therapy (ART) may help engage HIV-1-infected young people facing lifelong treatment. BREATHER showed short cycle therapy (SCT; 5 days on, 2 days off ART) was non-inferior to continuous therapy (CT) over 48 weeks. Planned follow-up was extended to 144 weeks, maintaining original randomisation. Methods: BREATHER was an open-label, non-inferiority trial. Participants aged 8-24yrs with virological suppression on efavirenz-based first-line ART were randomised 1:1, stratified by age and African/non-African sites, to remain on CT or change to SCT. The Kaplan-Meier method was used to estimate the proportion of participants with viral rebound (confirmed VL≥50 copies/mL) under intent-to-treat at 48 weeks (primary outcome), and in extended follow-up at 96, 144, and 192 weeks. SCT participants returned to CT following viral rebound, 3 VL blips or discontinuation of efavirenz. Findings: Of 199 participants (99 SCT, 100 CT), 97 per arm consented to extended follow-up. Median follow-up was 185.3 weeks (IQR 160.9–216.1). 69 (70%) SCT participants remained on SCT at last follow-up. 105 (53%) were male, baseline median age 14 years (IQR 12–18), median CD4 count 735 cells/μL (IQR 576–968). 16 SCT and 16 CT participants had confirmed VL≥50 copies/mL by the end of extended follow-up (HR 1.00, 95% CI 0.50–2.00). Estimated difference in percentage with viral rebound (SCT minus CT) by week 144 was 1.9% (90% CI -6.6–10.4; p = 0.72) and was similar in a per-protocol analysis. There were no significant differences between arms in proportions of participants with grade 3/4 adverse events (18 SCT vs 16 CT participants; p = 0.71) or ART-related adverse events (10 vs 12; p = 0.82). 20 versus 8 serious adverse events (SAEs) were reported in 16 SCT versus 4 CT participants, respectively (p = 0.005 comparing proportions between groups; incidence rate ratio 2.49, 95%CI 0.71–8.66, p = 0.15). 75% of SAEs (15 SCT, 6 CT) were hospitalisations for a wide range of conditions. 3 SCT and 6 CT participants switched to second-line ART following viral failure (p = 0.50). Conclusions: Sustainable non-inferiority of virological suppression in young people was shown for SCT versus CT over median 3.6 years. Standard-dose efavirenz-based SCT is a viable option for virologically suppressed HIV-1 infected young people on first-line ART with 3-monthly VL monitoring. Trial registration: EudraCT 2009-012947-40